Wan Eny Zarina Wan Abdul Rahman

Comparison of Balloon Snake and GVF Snake in Segmenting Masses from Breast Ultrasound Images

The active contour or Snake is a computational model that can trace boundaries of images. It is formulated based on controlled continuous splines and adopts energy minimization conception. This paper presents the application of Snakes for the segmentation of masses on breast ultrasound images. The boundaries of the masses identified may be used in locating potential cancerous cases for further analysis. Initially, Balloon Snake and Gradient Vector Flow (GVF) Snake are applied in segmenting the masses in the breast ultrasound phantom images. Comparison on the masses areas segmented by the Balloon Snake and the GVF Snake is done against the actual masses area. The better method with smaller values of average percentage area difference is chosen to be applied in segmenting masses on real breast ultrasound images. The performance is measured in terms of average percentage area difference traced by the chosen method against the area traced by expert radiologist. It is found that from eighty breast ultrasound phantom images (40 cyst and 40 solid masses) tested, the values of average percentage area difference for cyst and solid masses in the Balloon Snake are 3.07% and 9.17% respectively while GVF Snake are 13.43% and 48.37% respectively. Therefore Balloon Snake is chosen to segment real breast ultrasound images. Fifty images are tested. Segmentation on the images shows that the average percentage area difference of Balloon Snake is 4.29% which mean 95.71% accurate.

Seed point selection for seed-based region growing in segmenting microcalcifications

Seed-based region growing (SBRG) has been widely used as a segmentation method for medical images. The selection of initial seed point in SBRG is the crucial part before the segmentation process is carried out. Most of the region growing methods identify the seed point manually which involve human interaction and require prior information about the image. In this paper, an automated initial seed point selection for SBRG algorithm is proposed. The proposed method is tested on 50 mammogram images confirmed by a radiologist to consist microcalcifications. The performance is evaluated using Receiving Operator Curve (ROC) based on level of detection. Experimental results show that the method has successfully segmented the microcalcifications with 0.98 accuracy.

Improvement of digital mammogram images using histogram equalization, histogram stretching and median filter

Breast cancer is one of the most important diseases in females worldwide. According to the Malaysian Oncological Society, about 4% of women who are 40 years old and above are involved have breast cancer. Masses and microcalcifications are two important signs of breast cancer diagnosis on mammography. Enhancement techniques, i.e. histogram equalization, histogram stretching and median filters, were used to provide better visualization for radiologists in order to help early detection of breast abnormalities. In this research 60 digital mammogram images which includes 20 normal and 40 confirmed diagnosed cancerous cases were selected and manipulated using the mentioned techniques. The original and manipulated images were scored by three expert radiologists. Results showed that the selected methods have a positive significant effect on image quality.

A Pilot Study In Image Enhancement In Computed Radiography Mammogram Images Using Histogram Stretching Method

Efforts on the development of Computer Aided Diagnosis (CADG) system have been extensive. However, locally made CADG systems suitable for Malaysian patients are still unavailable. Many present CADG systems are developed based on Caucasian data. These systems may not be suitable for Malaysian data because the breast tissues of Malaysian women may differ from the Caucasians. This project initiates a study to develop a CADG system suitable for local data. This paper reports the image enhancement stage of 42 computed radiography mammogram images collected from the National Cancer Society of Malaysia. The method of histogram stretching is used to enhance these images which are then evaluated by radiologists based on the BIRADS standardization.

Classification of breast abnormalities using artificial neural network

Classification is the process of recognition, differentiation and categorizing objects into groups. Breast abnormalities are calcifications which are tumor markers that indicate the presence of cancer in the breast. The aims of this research are to classify the types of breast abnormalities using artificial neural network (ANN) classifier and to evaluate the accuracy performance using receiver operating characteristics (ROC) curve. The methods used in this research are ANN for breast abnormalities classifications and Canny edge detector as a feature extraction method. Previously the ANN classifier provides only the number of benign and malignant cases without providing information for specific cases. However in this research, the type of abnormality for each image can be obtained. The existing MIAS MiniMammographic database classified the mammogram images into three features only namely characteristic of background tissues, class of abnormality and radius of abnormality. However, in this research three ot...

Performance comparison of Canny and Sobel edge detectors on Balloon Snake in segmenting masses

The most common problem in image processing is image segmentation. One of the methods which can segment an image with a high accuracy is Balloon Snake. It uses the energy minimization conception where it has a dynamic behavior that deforms from an initial position and converges to the boundary of the object in an image. However, we found that the accurateness always influences by the edge map detector that being used in implementing the Balloon Snake. Edge map detector is used to strengthen the boundary of the object before the object is segmented. Two popular edge map detectors are chosen in this research namely Canny and Sobel edge detector. Both edge detectors are implemented in Balloon Snake in segmenting 40 masses in breast ultrasound images. The pixel area traced by the two combination methods namely Canny and Balloon Snake and Sobel and Balloon Snake are evaluated. The accuracy is measured based on the percentage pixel area difference between radiologist and the both combination methods. It is found that the combination of using Canny and Balloon Snake give 3.5% percentage of pixel area difference which is smaller compared to the other combination which have 12.7% percentage pixel area difference.

Review methods for image segmentation from computed tomography images

Image segmentation is a challenging process in order to get the accuracy of segmentation, automation and robustness especially in medical images. There exist many segmentation methods that can be implemented to medical images but not all methods are suitable. For the medical purposes, the aims of image segmentation are to study the anatomical structure, identify the region of interest, measure tissue volume to measure growth of tumor and help in treatment planning prior to radiation therapy. In this paper, we present a review method for segmentation purposes using Computed Tomography (CT) images. CT images has their own characteristics that affect the ability to visualize anatomic structures and pathologic features such as blurring of the image and visual noise. The details about the methods, the goodness and the problem incurred in the methods will be defined and explained. It is necessary to know the suitable segmentation method in order to get accurate segmentation. This paper can be a guide to researcher to choose the suitable segmentation method especially in segmenting the images from CT scan.